Long read assembler comparison

Load libraries

library(tidyverse)

## ── Attaching packages ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────── tidyverse 1.2.1 ──

## ✔ ggplot2 3.1.0       ✔ purrr   0.3.0  
## ✔ tibble  2.0.1       ✔ dplyr   0.8.0.1
## ✔ tidyr   0.8.2       ✔ stringr 1.4.0  
## ✔ readr   1.3.1       ✔ forcats 0.4.0

## ── Conflicts ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()

library(scales)

## 
## Attaching package: 'scales'

## The following object is masked from 'package:purrr':
## 
##     discard

## The following object is masked from 'package:readr':
## 
##     col_factor

library(tools)
library(gridExtra)

## 
## Attaching package: 'gridExtra'

## The following object is masked from 'package:dplyr':
## 
##     combine

library(ggforce)
library(knitr)
library(plyr)

## -------------------------------------------------------------------------

## You have loaded plyr after dplyr - this is likely to cause problems.
## If you need functions from both plyr and dplyr, please load plyr first, then dplyr:
## library(plyr); library(dplyr)

## -------------------------------------------------------------------------

## 
## Attaching package: 'plyr'

## The following objects are masked from 'package:dplyr':
## 
##     arrange, count, desc, failwith, id, mutate, rename, summarise,
##     summarize

## The following object is masked from 'package:purrr':
## 
##     compact

library(dplyr)
opts_chunk$set(dpi=300, fig.path='./plots/', echo=T, dev=c('png','pdf'), warning=FALSE, message=FALSE)

Load in the data

Problem set data:

adapter_length_results <- read_tsv("adapter_length_results") %>% filter(contigs > 0)
chimera_results <- read_tsv("chimera_results") %>% filter(contigs > 0)
glitch_results <- read_tsv("glitch_results") %>% filter(contigs > 0)
read_depth_results <- read_tsv("read_depth_results") %>% filter(contigs > 0)
read_identity_results <- read_tsv("read_identity_results") %>% filter(contigs > 0)
read_length_results <- read_tsv("read_length_results") %>% filter(contigs > 0)
random_junk_results <- read_tsv("random_junk_results") %>% filter(contigs > 0)
glitch_levels <- read_tsv("glitch_levels")

Simulated chromosome data:

sim_chromosome_results <- read_tsv("good_set_results") %>% filter(contiguity > 0.0)  # drop any that were total failures

Real chromosome data:

real_chromosome_results <- read_tsv("real_set_chromosome_results") %>% filter(contiguity > 0.0) %>% filter(assembler != "flye_v2.4.2_plasmids") %>% filter(assembler != "flye_v2.4.2_meta") %>% filter(assembler != "flye_v2.4.2_plasmids_meta")
real_chromosome_results$platform <- factor(real_chromosome_results$platform, levels = c("pacbio", "nanopore"))
real_chromosome_results$platform_int <- (real_chromosome_results$platform == "nanopore") + 16L

Simulated plasmid data:

sim_plasmid_results <- read_tsv("plasmid_results") %>% filter(name != 'chromosome') %>% filter(depth >= 1.0) %>% filter(depth <= 1000.0)
sim_plasmid_results_chromosome <- read_tsv("plasmid_results") %>% filter(name == 'chromosome')

Real plasmid data:

real_plasmid_results <- read_tsv("real_set_all_replicon_results") %>% filter(name != "1") %>% filter(depth > 5.0)

Combine plasmid data:

sim_plasmid_results$type <- "simulated"
real_plasmid_results$type <- "real"
plasmid_results <- rbind(sim_plasmid_results, real_plasmid_results)
plasmid_results$shape <- paste(plasmid_results$type, plasmid_results$contig_match, sep = "_")
plasmid_results$shape <- factor(plasmid_results$shape, levels = c("simulated_FALSE", "simulated_TRUE", "real_FALSE", "real_TRUE"))

Problem set plots

Functions for axes transformations and data prep:

contigs_to_str <- function(x) {
  if (x == 1) {
    return("1")
  } else {
    return("2+")
  }
}

# These describe the y-axis transformation. I used a sigmoid function to focus on the range near 100%.
y_trans <- function(x) {
  a <- exp((x-1)/0.15)
  return(a/(a+1))
}
y_trans_inv <- function(x) {
  return(0.15 * log(-x/(x-1)) + 1)
}
trans_sigmoid <- trans_new(name = "sigmoid", transform = y_trans, inverse = y_trans_inv)

prep_results_one_assembler <- function(results, assembler_name) {
  if (is.null(assembler_name)) { 
    assembler_results <- results
  } else {
    assembler_results <- results[results$assembler == assembler_name,]
  }
  assembler_results$contigs_str <- sapply(assembler_results$contigs, contigs_to_str)
  return(assembler_results)
}

Functions for making the plots:

plot_adapter_length_one_assembler <- function(results, assembler_name, colour) {
  assembler_results <- prep_results_one_assembler(results, assembler_name)
  p <- ggplot(data=assembler_results) +
    geom_point(aes(x=start_adapter_length, y=contiguity, shape=contigs_str), colour=colour, alpha=0.75) +
    scale_shape_manual("Contigs", values = c(16, 1)) +
    coord_trans(y = trans_sigmoid, limy=c(0.0, 1.05)) +
    scale_y_continuous(expand=c(0.02, 0), limits=c(0.0, 1.05),
                       breaks = seq(0.7, 1.0, 0.1),
                       minor_breaks = seq(0, 0.6, 0.1),
                       labels = scales::percent_format(accuracy = 1)) +
    scale_x_continuous(expand=c(0, 0), limits=c(0, 1000)) +
    theme_bw() + theme(panel.grid.minor = element_line(size = 0.5), panel.grid.major = element_line(size = 0.5)) +
    labs(title = NULL, x = "Adapter length", y = "Contiguity")
  return(p)
}

plot_chimera_rate_one_assembler <- function(results, assembler_name, colour) {
  assembler_results <- prep_results_one_assembler(results, assembler_name)
  p <- ggplot(data=assembler_results) +
    geom_point(aes(x=chimera_rate, y=contiguity, shape=contigs_str), colour=colour, alpha=0.75) +
    scale_shape_manual("Contigs", values = c(16, 1)) +
    coord_trans(y = trans_sigmoid, limy=c(0.0, 1.05)) +
    scale_y_continuous(expand=c(0.02, 0), limits=c(0.0, 1.05),
                       breaks = seq(0.7, 1.0, 0.1),
                       minor_breaks = seq(0, 0.6, 0.1),
                       labels = scales::percent_format(accuracy = 1)) +
    scale_x_continuous(expand=c(0, 0), limits=c(0.0, 0.25), labels = scales::percent_format(accuracy = 1)) +
    theme_bw() + theme(panel.grid.minor = element_line(size = 0.5), panel.grid.major = element_line(size = 0.5)) +
    labs(title = NULL, x = "Chimera rate", y = "Contiguity")
  return(p)
}

plot_glitch_level_one_assembler <- function(results, assembler_name, colour) {
  assembler_results <- prep_results_one_assembler(results, assembler_name)
  p <- ggplot(data=assembler_results) +
    geom_point(aes(x=glitch_size, y=contiguity, shape=contigs_str), colour=colour, alpha=0.75) +
    scale_shape_manual("Contigs", values = c(16, 1)) +
    coord_trans(y = trans_sigmoid, limy=c(0.0, 1.05)) +
    scale_y_continuous(expand=c(0.02, 0), limits=c(0.0, 1.05),
                       breaks = seq(0.7, 1.0, 0.1),
                       minor_breaks = seq(0, 0.6, 0.1),
                       labels = scales::percent_format(accuracy = 1)) +
    scale_x_continuous(expand=c(0, 0), limits=c(0, 100)) +
    theme_bw() + theme(panel.grid.minor = element_line(size = 0.5), panel.grid.major = element_line(size = 0.5)) +
    labs(title = NULL, x = "Glitch level", y = "Contiguity")
  return(p)
}

plot_read_depth_one_assembler <- function(results, assembler_name, colour) {
  assembler_results <- prep_results_one_assembler(results, assembler_name)
  p <- ggplot(data=assembler_results) +
    geom_point(aes(x=depth, y=contiguity, shape=contigs_str), colour=colour, alpha=0.75) +
    scale_shape_manual("Contigs", values = c(16, 1)) +
    coord_trans(y = trans_sigmoid, limy=c(0.0, 1.05)) +
    scale_y_continuous(expand=c(0.02, 0), limits=c(0.0, 1.05),
                       breaks = seq(0.7, 1.0, 0.1),
                       minor_breaks = seq(0, 0.6, 0.1),
                       labels = scales::percent_format(accuracy = 1)) +
    scale_x_continuous(expand=c(0, 0), limits=c(0, 100)) +
    theme_bw() + theme(panel.grid.minor = element_line(size = 0.5), panel.grid.major = element_line(size = 0.5)) +
    labs(title = NULL, x = "Read depth", y = "Contiguity")
  return(p)
}

plot_read_identity_one_assembler <- function(results, assembler_name, colour) {
  assembler_results <- prep_results_one_assembler(results, assembler_name)
  p <- ggplot(data=assembler_results) +
    geom_point(aes(x=read_identity_mean, y=contiguity, shape=contigs_str), colour=colour, alpha=0.75) +
    scale_shape_manual("Contigs", values = c(16, 1)) +
    coord_trans(y = trans_sigmoid, limy=c(0.0, 1.05)) +
    scale_y_continuous(expand=c(0.02, 0), limits=c(0.0, 1.05),
                       breaks = seq(0.7, 1.0, 0.1),
                       minor_breaks = seq(0, 0.6, 0.1),
                       labels = scales::percent_format(accuracy = 1)) +
    scale_x_continuous(expand=c(0, 0), limits=c(0.7, 1.0), labels = scales::percent_format(accuracy = 1)) +
    theme_bw() + theme(panel.grid.minor = element_line(size = 0.5), panel.grid.major = element_line(size = 0.5)) +
    labs(title = NULL, x = "Read identity", y = "Contiguity")
  return(p)
}

plot_read_length_one_assembler <- function(results, assembler_name, colour) {
  assembler_results <- prep_results_one_assembler(results, assembler_name)
  p <- ggplot(data=assembler_results) +
    geom_point(aes(x=fragment_length_mean, y=contiguity, shape=contigs_str), colour=colour, alpha=0.75) +
    scale_shape_manual("Contigs", values = c(16, 1)) +
    coord_trans(y = trans_sigmoid, limy=c(0.0, 1.05)) +
    scale_y_continuous(expand=c(0.02, 0), limits=c(0.0, 1.05),
                       breaks = seq(0.7, 1.0, 0.1),
                       minor_breaks = seq(0, 0.6, 0.1),
                       labels = scales::percent_format(accuracy = 1)) +
    scale_x_continuous(expand=c(0, 0), limits=c(0, 50000)) +
    theme_bw() + theme(panel.grid.minor = element_line(size = 0.5), panel.grid.major = element_line(size = 0.5)) +
    labs(title = NULL, x = "Read length", y = "Contiguity")
  return(p)
}

plot_random_junk_one_assembler <- function(results, assembler_name, colour) {
  assembler_results <- prep_results_one_assembler(results, assembler_name)
  p <- ggplot(data=assembler_results) +
    geom_point(aes(x=junk_rate, y=contiguity, shape=contigs_str), colour=colour, alpha=0.75) +
    scale_shape_manual("Contigs", values = c(16, 1)) +
    coord_trans(y = trans_sigmoid, limy=c(0.0, 1.05)) +
    scale_y_continuous(expand=c(0.02, 0), limits=c(0.0, 1.05),
                       breaks = seq(0.7, 1.0, 0.1),
                       minor_breaks = seq(0, 0.6, 0.1),
                       labels = scales::percent_format(accuracy = 1)) +
    scale_x_continuous(expand=c(0, 0), limits=c(0.0, 0.25), labels = scales::percent_format(accuracy = 1)) +
    theme_bw() + theme(panel.grid.minor = element_line(size = 0.5), panel.grid.major = element_line(size = 0.5)) +
    labs(title = NULL, x = "Random/junk rate", y = "Contiguity")
  return(p)
}

p1 <- plot_adapter_length_one_assembler(adapter_length_results, "canu_v1.8", "#AB3B33")
p2 <- plot_adapter_length_one_assembler(adapter_length_results, "flye_v2.4.2", "#998B00")
p3 <- plot_adapter_length_one_assembler(adapter_length_results, "ra_07364a1", "#009E43")
p4 <- plot_adapter_length_one_assembler(adapter_length_results, "unicycler_v0.4.7", "#006DC4")
p5 <- plot_adapter_length_one_assembler(adapter_length_results, "wtdbg2_v2.4", "#AC50B5")
grid.arrange(p1, p2, p3, p4, p5, ncol = 1)

p1 <- plot_chimera_rate_one_assembler(chimera_results, "canu_v1.8", "#AB3B33")
p2 <- plot_chimera_rate_one_assembler(chimera_results, "flye_v2.4.2", "#998B00")
p3 <- plot_chimera_rate_one_assembler(chimera_results, "ra_07364a1", "#009E43")
p4 <- plot_chimera_rate_one_assembler(chimera_results, "unicycler_v0.4.7", "#006DC4")
p5 <- plot_chimera_rate_one_assembler(chimera_results, "wtdbg2_v2.4", "#AC50B5")
grid.arrange(p1, p2, p3, p4, p5, ncol = 1)

p1 <- plot_glitch_level_one_assembler(glitch_results, "canu_v1.8", "#AB3B33")
p2 <- plot_glitch_level_one_assembler(glitch_results, "flye_v2.4.2", "#998B00")
p3 <- plot_glitch_level_one_assembler(glitch_results, "ra_07364a1", "#009E43")
p4 <- plot_glitch_level_one_assembler(glitch_results, "unicycler_v0.4.7", "#006DC4")
p5 <- plot_glitch_level_one_assembler(glitch_results, "wtdbg2_v2.4", "#AC50B5")
grid.arrange(p1, p2, p3, p4, p5, ncol = 1)

p1 <- plot_read_depth_one_assembler(read_depth_results, "canu_v1.8", "#AB3B33")
p2 <- plot_read_depth_one_assembler(read_depth_results, "flye_v2.4.2", "#998B00")
p3 <- plot_read_depth_one_assembler(read_depth_results, "ra_07364a1", "#009E43")
p4 <- plot_read_depth_one_assembler(read_depth_results, "unicycler_v0.4.7", "#006DC4")
p5 <- plot_read_depth_one_assembler(read_depth_results, "wtdbg2_v2.4", "#AC50B5")
grid.arrange(p1, p2, p3, p4, p5, ncol = 1)

p1 <- plot_read_identity_one_assembler(read_identity_results, "canu_v1.8", "#AB3B33")
p2 <- plot_read_identity_one_assembler(read_identity_results, "flye_v2.4.2", "#998B00")
p3 <- plot_read_identity_one_assembler(read_identity_results, "ra_07364a1", "#009E43")
p4 <- plot_read_identity_one_assembler(read_identity_results, "unicycler_v0.4.7", "#006DC4")
p5 <- plot_read_identity_one_assembler(read_identity_results, "wtdbg2_v2.4", "#AC50B5")
grid.arrange(p1, p2, p3, p4, p5, ncol = 1)

p1 <- plot_read_length_one_assembler(read_length_results, "canu_v1.8", "#AB3B33")
p2 <- plot_read_length_one_assembler(read_length_results, "flye_v2.4.2", "#998B00")
p3 <- plot_read_length_one_assembler(read_length_results, "ra_07364a1", "#009E43")
p4 <- plot_read_length_one_assembler(read_length_results, "unicycler_v0.4.7", "#006DC4")
p5 <- plot_read_length_one_assembler(read_length_results, "wtdbg2_v2.4", "#AC50B5")
grid.arrange(p1, p2, p3, p4, p5, ncol = 1)

p1 <- plot_random_junk_one_assembler(random_junk_results, "canu_v1.8", "#AB3B33")
p2 <- plot_random_junk_one_assembler(random_junk_results, "flye_v2.4.2", "#998B00")
p3 <- plot_random_junk_one_assembler(random_junk_results, "ra_07364a1", "#009E43")
p4 <- plot_random_junk_one_assembler(random_junk_results, "unicycler_v0.4.7", "#006DC4")
p5 <- plot_random_junk_one_assembler(random_junk_results, "wtdbg2_v2.4", "#AC50B5")
grid.arrange(p1, p2, p3, p4, p5, ncol = 1)

I also make this little reference plot for the glitch levels:

p1 <- ggplot(data=glitch_levels) +
  geom_line(aes(x=level, y=size), colour="#f8766d", size=1) +
  scale_y_continuous(minor_breaks = NULL) +
  scale_x_continuous(minor_breaks = NULL) +
  theme_bw() +
  labs(x = "Glitch level", y = "Glitch size/skip")
p2 <- ggplot(data=glitch_levels) +
  geom_line(aes(x=level, y=distance), colour="#619cff", size=1) +
  scale_y_log10(labels = comma, breaks = c(100, 1000, 10000, 100000), minor_breaks = NULL) +
  scale_x_continuous(minor_breaks = NULL) +
  theme_bw() +
  labs(x = "Glitch level", y = "Glitch rate")



p1 <- ggplotGrob(p1)
p2 <- ggplotGrob(p2)
maxWidth = grid::unit.pmax(p1$widths[2:5], p2$widths[2:5])
p1$widths[2:5] <- as.list(maxWidth)
p2$widths[2:5] <- as.list(maxWidth)
grid.arrange(p1, p2, ncol = 2)

Contiguity plots

Functions for axes transformations and data prep:

contigs_to_str <- function(x) {
  if (x == 1) {
    return("1")
  } else {
    return("2+")
  }
}

# These describe the contiguity y-axis transformation. I used a sigmoid function to focus on the range near 100%.
y_trans <- function(x) {
  a <- exp((x-1)/0.1)
  return(a/(a+1))
}
y_trans_inv <- function(x) {
  return(0.1 * log(-x/(x-1)) + 1)
}
trans_sigmoid <- trans_new(name = "sigmoid", transform = y_trans, inverse = y_trans_inv)

# These describe the identity y-axis transformation.
id_y_trans <- function(x) {
  return(-10.0 * log10(1-x))
}
id_y_trans_inv <- function(x) {
  return(1.0 - (10.0 ** (x/-10.0)))
}
trans_phred <- trans_new(name = "phred", transform = id_y_trans, inverse = id_y_trans_inv)

sim_chromosome_results$contiguity_trans <- y_trans(sim_chromosome_results$contiguity)
p1 <- ggplot(data=sim_chromosome_results) +
  geom_sina(aes(x=assembler, y=contiguity_trans, colour=assembler), scale="area", maxwidth=0.85, bw=0.02, size=1, alpha=0.75, stroke=0) +
  scale_colour_manual(values = c("canu_v1.8" = "#AB3B33", "flye_v2.4.2" = "#998B00", "ra_07364a1" = "#009E43", "unicycler_v0.4.7" = "#006DC4", "wtdbg2_v2.4" = "#AC50B5"), guide=FALSE) +
  theme_bw() + theme(panel.grid.minor = element_line(size = 0.5), panel.grid.major = element_line(size = 0.5)) +
  scale_y_continuous(expand=c(0.01, 0.0),
                     breaks = y_trans(c(0.0, 0.7, 0.8, 0.9, 1.0)),
                     labels = c("0%", "70%", "80%", "90%", "100%"),
                     minor_breaks = y_trans(seq(0.1, 0.6, 0.1))) +
  scale_x_discrete(labels = gsub("_", " ", c("Canu\nv1.8", "Flye\nv2.4.2", "Ra\n07364a1", "Unicycler\nv0.4.7", "Wtdbg2\nv2.4"))) +
  coord_cartesian(ylim=c(0, 0.6)) +
  labs(title="Simulated read sets", y="Contiguity", x=NULL)

p2 <- ggplot(data=sim_chromosome_results) +
  geom_sina(aes(x=assembler, y=contiguity, colour=assembler), scale="area", maxwidth=0.85, bw=0.000005, size=1, alpha=0.75, stroke=0) +
  scale_colour_manual(values = c("canu_v1.8" = "#AB3B33", "flye_v2.4.2" = "#998B00", "ra_07364a1" = "#009E43", "unicycler_v0.4.7" = "#006DC4", "wtdbg2_v2.4" = "#AC50B5"), guide=FALSE) +
  theme_bw() + theme(panel.grid.minor = element_line(size = 0.5), panel.grid.major = element_line(size = 0.5)) +
  scale_y_continuous(labels = scales::percent_format(accuracy=0.001), limits = c(0.9999, 1.0001)) +
  scale_x_discrete(labels = gsub("_", " ", c("Canu\nv1.8", "Flye\nv2.4.2", "Ra\n07364a1", "Unicycler\nv0.4.7", "Wtdbg2\nv2.4"))) +
  coord_cartesian(ylim = c(0.99992, 1.00008)) +
  labs(title=" ", y=NULL, x=NULL)

real_chromosome_results$contiguity_trans <- y_trans(real_chromosome_results$contiguity)
p3 <- ggplot(data=real_chromosome_results) +
  geom_sina(aes(x=assembler, y=contiguity_trans, colour=assembler, shape=platform_int), scale="area", maxwidth=0.85, bw=0.02, size=1.75, alpha=0.75, stroke=0) +
  scale_shape_identity("Platform", labels=c("PacBio RSII", "ONT MinION")) +
  scale_colour_manual(values = c("canu_v1.8" = "#AB3B33", "flye_v2.4.2" = "#998B00", "ra_07364a1" = "#009E43", "unicycler_v0.4.7" = "#006DC4", "wtdbg2_v2.4" = "#AC50B5"), guide=FALSE) +
  theme_bw() + theme(panel.grid.minor = element_line(size = 0.5), panel.grid.major = element_line(size = 0.5)) +
  scale_y_continuous(expand=c(0.01, 0.0),
                     breaks = y_trans(c(0.0, 0.7, 0.8, 0.9, 1.0)),
                     labels = c("0%", "70%", "80%", "90%", "100%"),
                     minor_breaks = y_trans(seq(0.1, 0.6, 0.1))) +
  scale_x_discrete(labels = gsub("_", " ", c("Canu\nv1.8", "Flye\nv2.4.2", "Ra\n07364a1", "Unicycler\nv0.4.7", "Wtdbg2\nv2.4"))) +
  coord_cartesian(ylim=c(0, 0.6)) +
  labs(title="Real read sets", y="Contiguity", x=NULL)

p4 <- ggplot(data=real_chromosome_results) +
  geom_sina(aes(x=assembler, y=contiguity, colour=assembler, shape=platform_int), scale="area", maxwidth=0.85, bw=0.000005, size=1.75, alpha=0.75, stroke=0) +
  scale_shape_identity("Platform", labels=c("PacBio RSII", "ONT MinION")) +
  scale_colour_manual(values = c("canu_v1.8" = "#AB3B33", "flye_v2.4.2" = "#998B00", "flye_v2.4.2_plasmids" = "#998B00", "flye_v2.4.2_meta" = "#998B00", "flye_v2.4.2_plasmids_meta" = "#998B00", "ra_07364a1" = "#009E43", "unicycler_v0.4.7" = "#006DC4", "wtdbg2_v2.4" = "#AC50B5"), guide=FALSE) +
  theme_bw() + theme(panel.grid.minor = element_line(size = 0.5), panel.grid.major = element_line(size = 0.5)) +
  scale_y_continuous(labels = scales::percent_format(accuracy=0.001), limits = c(0.9999, 1.0001)) +
  scale_x_discrete(labels = gsub("_", " ", c("Canu\nv1.8", "Flye\nv2.4.2", "Ra\n07364a1", "Unicycler\nv0.4.7", "Wtdbg2\nv2.4"))) +
  coord_cartesian(ylim = c(0.99992, 1.00008)) +
  labs(title=" ", y=NULL, x=NULL)

p1 <- ggplotGrob(p1)
p2 <- ggplotGrob(p2)
p3 <- ggplotGrob(p3)
p4 <- ggplotGrob(p4)
maxWidth = grid::unit.pmax(p1$widths[2:5], p2$widths[2:5], p3$widths[2:5], p4$widths[2:5])
p1$widths[2:5] <- as.list(maxWidth)
p2$widths[2:5] <- as.list(maxWidth)
p3$widths[2:5] <- as.list(maxWidth)
p4$widths[2:5] <- as.list(maxWidth)
grid.arrange(p1, p2, p3, p4, ncol = 2)

Identity plots

real_chromosome_results$identity_trans <- id_y_trans(real_chromosome_results$identity)
ggplot(data=real_chromosome_results) +
  geom_sina(aes(x=assembler, y=identity_trans, colour=assembler, shape=platform_int), scale="area", maxwidth=0.85, bw=1, size=1.75, alpha=0.75, stroke=0) +
  scale_shape_identity("Platform", labels=c("PacBio RSII", "ONT MinION")) +
  scale_colour_manual(values = c("canu_v1.8" = "#AB3B33", "flye_v2.4.2" = "#998B00", "ra_07364a1" = "#009E43", "unicycler_v0.4.7" = "#006DC4", "wtdbg2_v2.4" = "#AC50B5"), guide=FALSE) +
  coord_cartesian(ylim=c(10, 56)) +
  scale_y_continuous(breaks = seq(10, 50, 10), minor_breaks = NULL) +
  scale_x_discrete(labels = gsub("_", " ", c("Canu\nv1.8", "Flye\nv2.4.2", "Ra\n07364a1", "Unicycler\nv0.4.7", "Wtdbg2\nv2.4"))) +
  theme_bw() +
  labs(title="Mean sequence identity", x=NULL, y="Qscore")

Time plots

ggplot(data=real_chromosome_results) +
  geom_sina(aes(x=assembler, y=minutes, colour=assembler, shape=platform_int), scale="area", maxwidth=0.85, bw=0.12, size=1.75, alpha=0.75, stroke=0) +
  scale_shape_identity("Platform", labels=c("PacBio RSII", "ONT MinION")) +
  scale_colour_manual(values = c("canu_v1.8" = "#AB3B33", "flye_v2.4.2" = "#998B00", "flye_v2.4.2_plasmids" = "#998B00", "flye_v2.4.2_meta" = "#998B00", "flye_v2.4.2_plasmids_meta" = "#998B00", "ra_07364a1" = "#009E43", "unicycler_v0.4.7" = "#006DC4", "wtdbg2_v2.4" = "#AC50B5"), guide=FALSE) +
  scale_y_log10(breaks = c(1, 3, 10, 30, 100, 300, 1000, 3000), minor_breaks = NULL, limits = c(2.9, 1430)) +
  scale_x_discrete(labels = gsub("_", " ", c("Canu\nv1.8", "Flye\nv2.4.2", "Ra\n07364a1", "Unicycler\nv0.4.7", "Wtdbg2\nv2.4"))) +
  theme_bw() +
  labs(title="Assembly time", y="Time (minutes)", x=NULL)

Making a smaller version of the time plot for my poster showing only the ONT results:

nanopore_chromosome_results <- filter(real_chromosome_results, platform == "nanopore")
ggplot(data=nanopore_chromosome_results) +
  geom_sina(aes(x=assembler, y=minutes, colour=assembler), scale="area", maxwidth=0.85, bw=0.15, size=2.75, alpha=0.75, stroke=0) +
  scale_colour_manual(values = c("canu_v1.8" = "#AB3B33", "flye_v2.4.2" = "#998B00", "flye_v2.4.2_plasmids" = "#998B00", "flye_v2.4.2_meta" = "#998B00", "flye_v2.4.2_plasmids_meta" = "#998B00", "ra_07364a1" = "#009E43", "unicycler_v0.4.7" = "#006DC4", "wtdbg2_v2.4" = "#AC50B5"), guide=FALSE) +
  scale_y_log10(breaks = c(1, 3, 10, 30, 100, 300, 1000, 3000), minor_breaks = NULL, limits = c(2.9, 1430)) +
  scale_x_discrete(labels = gsub("_", " ", c("Canu\nv1.8", "Flye\nv2.4.2", "Ra\n07364a1", "Unicycler\nv0.4.7", "Wtdbg2\nv2.4"))) +
  theme_bw() +
  labs(title="Assembly time", y="Time (minutes)", x=NULL)

Plasmid plots